Signals rT SAIN SL A calom oe Signals and Systems Simon Hayhkin McMaster University Barry Van Veen University of Wisconsin JOHN WILEY & Sons, INc. New York @ Chichester # Weinheim # Brisbane @ Singapore 8 TorontoTo Nancy and Kathy, Emily, David, and Jonathan EDITOR Bill Zobrist MARKETING MANAGER Katherine Hepburn SENIOR PRODUCTION MANAGER = Lucille Buonocore SENIOR PRODUCTION EDITOR = Monique Caleito SENIOR DESIGNER Laura Boucher TEXT DESIGNER Nancy Field COVER DESIGNER Laura Boucher COVER PHOTO) Courtesy of NASA ILLUSTRATION EDITOR = Sigmund Malinowski ILLUSTRATION = Wellington Studios This book was set in Times Roman by UG division of GGS Information Services and printed and bound by Quebecor Printing, Kingsport. The cover was printed by Phoenix Color Corporation. This book is printed on acid-free paper. The paper in this book was manufactured by a mill whose forest management programs include sustained yield harvesting of its timberlands. Sustained yield harvesting principles ensure that the numbers of trees cu cach year does not exceed the amount of new growth. Copyright © 1999, John Wiley & Sons, Inc. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or b any means, electronic, mechanical, photocopying, recording, scanning or otherwise, except as permitted unc Sections 107 or 108 of the 1976 United States Copyright Act, without either che prior written permission a! the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearanex Center, 222 Rosewood Drive, Danvers, MA 01923, (508) 750-8400, fax (508) 750-4470. Requests co the Publisher for permission shauld be addressed to the Permissions Department, Jahn Wiley & Sons, Inc., 605 Third Avenue, New York, NY 10158-0012, (212) 850-6011, fax (212) 850-6008, E-Mail: PERMREQ@WILEY.COM. Library of Congress Cataloging-in-Publication Data Haykin, Simon Signals and systems / Simon Haykin, Barry Van Veen. p. cm, Includes index. ISBN 0-471-13820-7 (cloth : alk. paper} 1. Signal processing. 2. System analysis. 3. Linear time Invariant systems. 4. Telecommunication systems. I. Van Veen, Barry. IL Title. TK5102.5.H37 1999 621.382'2—de21 97-52090 CIP Printed in the United States of America 109876Xvili CONTENTS T, settling time X(t, jar) short-time Fourier transform of x(t) W.(7, a) wavelet transform of x(t) Abbreviations A/D analog-to-digital (converter) AM amplitude modulation BIBO bounded input bounded output CW continuous wave D/A digital-to-analog (converter) dB decibel DOF degree of freedom DSB-SC double sideband-suppressed carrier DTFS discrete-time Fourier series DTFT discrete-time Fourier transform FDM __ frequency-division multiplexing FFT fast Fourier transform FIR finite-duration impulse response FM frequency modulation FS Fourier series FT Fourier transform Hz hertz IIR infinite-duration impulse response LT! linear time-invariant (system) MRI magnetic resonance image MSE mean squared error PAM _ pulse-amplitude modulation PCM __ pulse-code modulation PM phase modulation QAM — quadrature-amplitude modulation ROC _ region of convergence rad radian(s} S second SSB single sideband modulation STFT short-time Fourier transform TDM _ time-division multiplexing VSB vestigial sideband modulation WT wavelet transformPREFACE Each “Exploring Concepts with MATLAB” section is designed to instruct the s on the preper application of the relevant MATLAB commands and develop add insight into the concepts introduced in the chapter. Minimal previous exposure to MA is assumed. The MATLAB code for all the computations performed in the book, inc the last chapter, are available on the Wiley Web Site: http://www.wiley.com/college There are 10 chapters in the book, organized as follows: » Chapter 1 begins by motivating the reader as to what signals and systems a how they arise in communication systems, control systems, remote sensing, bi ical signal processing, and the auditory system. It then describes the different, of signals, defines certain elementary signals, and introduces the basic notic volved in the characterization of systems. » Chapter 2 presents a detailed treatment of time-domain representations of time-invariant {LTI) systems. It develops convolution from the representatior input signal as a superposition of impulses. The notions of causality, memor bility, and invertibility that were briefly introduced in Chapter 1 are then re in terms of the impulse response description for LT] systems. The steady-st. sponse of a LTI system to a sinusoidal input is used to introduce the conc frequency response. Differential- and difference-equation representations for time-invariant systems are also presented. Next, block diagram representatio LTI systems are introduced. The chapter finishes with a discussion of the variable description of LTI systems. » Chapter 3 deals with the Fourier representation of signals. In particular, the F representations of four fundamental classes of signals are thoroughly discusse unified manner: » Discrete-time periodic signals: the discrete-time Fourier series >» Continuous-time periodic signals: the Fourier series » Discrete-time nonperiodic signals: the discrete-time Fourier transform » Continuous-time nonperiodic signals: the Fourier transform A novel feature of the chapter is the way in which similarities between thes: representations are exploited and the differences between them are highlightec fact that complex sinusoids are eigenfunctions of LTI systems is used to motiva representation of signals in terms of complex sinusoids. The basic form of the Fi representation for each signal class is introduced and the four representatio: developed in sequence. Next, the properties of all four representations are st side by side. A strict separation between signal classes and the corresponding Fi representations is maintained throughout the chapter. It is our conviction t parallel, yet separate, treatment minimizes confusion between representation aids later mastery of proper application for each. Mixing of Fourier represent: occurs naturally in the context of analysis and computational applications < thus deterred to Chapter 4. » Chapter 4 presents a thorough treatment of the applications of Fourier repre: tions to the study of signals and LTI systems. Links between the frequency-dc and time-domain system representations presented in Chapter 2 are established. analysis and computational applications are then used to motivate derivation « relationships between the four Fourier representations and develop the student’ in applying these tools, The continuous-time and discrete-time Fourier trans representations of periodic signals are introduced for analyzing problems in there is a mixture of periodic and nonperiodic signals, such as application of riodic input to a LTI system. The Fourier transform representation for discretePreface vii signals is then developed as a tool for analyzing situations in which there is a mixture of continuous-time and discrete-time signals. The sampling process and continu- ous-time signal reconstruction from samples are studied in detail within this context. Systems for discrete-time processing of continuous-time signals are also discussed, including the issues of oversampling, decimation, and interpolation. The chapter concludes by developing relationships between the discrete-time Fourier series and the discrete-time and continuous-time Fourier transforms in order to introduce the computational aspects of the Fourier analysis of signals. » Chapter 5 presents an introductory treatment of linear modulation systems applied to communication systems. Practical reasons for using modulation are described. Amplitude modulation and its variants, namely, double sideband-suppressed carrier modulation, single sideband modulation, and vestigial sideband modulation, are dis- cussed. The chapter also includes a discussion of pulse-amplitude modulation and its role in digital communications to again highlight a natural interaction between continuous-time and discrete-time signals. The chapter includes a discussion of frequency-division and time-division multiplexing techniques. It finishes with a treat- ment of phase and group delays that arise when a modulated signal is transmitted through a linear channel. » Chapter 6 discusses the Laplace transform and its use for the complex exponential representations of continuous-time signals and the characterization of systems. The eigenfunction property of LTI systems and the existence of complex exponential representations for signals that have no Fourier representation are used to motivate the study of Laplace transforms. The unilateral Laplace transform is studied first and applied to the solution of differential equations with initial conditions to reflect the dominant role of the Laplace transform in engineering applications. The bilateral Laplace transform is introduced next and is used to study issues of causality, stability, invertibility, and the relationship between poles and zeros and frequency response. The relationships between the transfer function description of LTI systems and the time-domain descriptions introduced in Chapter 2 are developed. » Chapter 7 is devoted to the z-transform and its use in the complex exponential rep- resentation of discrete-time signals and the characterization of systems. As in Chapter 6, the z-transform is motivated as a more general representation than that of the discrete-time Fourier transform. Consistent with its primary role as an analysis tool, we begin with the bilateral z-transform. The properties of the z-transform and tech- niques for inversion are introduced. Next, the z-transform is used for transform analysis of systems. Relationships berween the transfer function and time-domain descriptions introduced in Chapter 2 are developed. Issues of invertibility, stability, causality, and the relationship berween the frequency response and poles and zeros are revisited. The use of the z-transform for deriving computational structures for implementing discrete-time systems on computers is introduced. Lastly, use of the unilateral z-transform for solving difference equations is presented. » Chapter 8 discusses the characterization and design of linear filters and equalizers. The approximation problem, with emphasis on Butterworth functions and brief men- tion of Chebyshev functions, is introduced. Direct and indirect methods for the design of analog (i.e., continuous-time) and digital (i.e., discrete-time) types of filters are presented. The window method for the design of finite-duration impulse response digital filters and the bilateral transform method for the design of infinite-duration impulse response digital filters are treated in detail. Filter design offers another op- portunity to reinforce the links between continuous-time and discrete-time systems. The chapter builds on material presented in Chapter 4 in developing a method for the